Abstract
There are two commonly used measures of boredom: the Boredom Proneness Scale (BPS) and the Boredom Susceptibility Scale (ZBS). Although both were designed to measure the propensity to experience boredom (i.e., trait boredom), there are reasons to think they may not measure the same construct. The present research sought to evaluate this proposition in several stages. Specifically, relationships between the BPS, ZBS, and important causal (Study 1, N = 837), correlational (Study 2, N = 233), and outcome variables (Study 3, N = 137) were examined in university students. Taken together, results support the notion that the BPS and ZBS do not measure the same construct. Specifically, higher BPS scores were associated with higher levels of neuroticism, experiential avoidance, attentional and nonplanning impulsivity, anxiety, depression, dysphoria, and emotional eating. Conversely, higher ZBS scores were associated with higher levels of motor impulsivity, sensitivity to reward, gambling, and alcohol use and lower levels of neuroticism, experiential avoidance, and sensitivity to punishment.
On first thought, boredom might seem like a trivial human experience—a temporary annoyance or dissatisfaction with the environment that can be easily remedied by engaging in something interesting. Although there may be some truth to this common conception, boredom is not always trivial. In fact, not only can boredom be an extreme and chronic experience for some, but research has also demonstrated its detrimental impact and association with psychopathology. For example, studies have shown that boredom is related to depression (Farmer & Sundberg, 1986), anxiety (Sommers & Vodanovich, 2000), psychotic disorders (Todman, 2003), alexithymia (Eastwood, Cavaliere, Fahlman, & Eastwood, 2007), lacking life meaning (Fahlman, Mercer, Gaskovski, Eastwood, & Eastwood, 2009), and increased mortality (Britton & Shipley, 2010). Boredom has also been associated with behavioral problems such as problem gambling (Blaszczynski, McConaghy, & Frankova, 1990), alcohol abuse (Wiesbeck et al., 1996), drug use (Paulson, Coombs, & Richardson, 1990), and binge eating (Stickney & Miltenberger, 1999). Despite these and other important associations, very little empirical attention has been devoted to developing psychometrically sound measures of boredom. In fact, in his review of existing boredom measures, Vodanovich (2003) noted that although research on boredom has increased in recent years, it is still scarce, particularly with regard to the measurement of boredom.
Several boredom measures have been developed, including, for example, the Job Boredom Scale (JBS; Lee, 1986), Boredom Coping Scale (BC; Hamilton, Haier, & Buchsbaum, 1984), Leisure Boredom Scale (LBS; Iso-Ahola & Weissinger, 1990), Free Time Boredom scale (FTB; Ragheb & Merydith, 2001), and the Sexual Boredom Scale (SBS; Watt & Ewing, 1996). Most of these scales, however, are limited in scope and have received little empirical attention (see Vodanovich, 2003, for a review). Two exceptions are the Boredom Proneness Scale (BPS), a 28-item self-report measure created by Farmer and Sundberg (1986), and the Boredom Susceptibility Scale (ZBS), a 10-item subscale of Zuckerman’s (1979) Sensation Seeking Scale, which, to date, are the most commonly used measures of boredom. Both of these scales were developed to measure the propensity to experience boredom (i.e., trait boredom); however, there are reasons to believe they may not measure the same specific boredom-related construct.
One reason to suggest that the BPS and ZBS measure something different is that they are not highly correlated. For example, in one study they were correlated at r = .25 (Farmer & Sundberg, 1986), and in another at r = .29 (Kass & Vodanovich, 1990). A possible reason for such weak relationships between the two scales is that the ZBS typically demonstrates poor reliability (i.e., coefficient alpha has been reported to range from .56 to .65; split-half reliability from .38 to .75; and 3-week test–retest reliability at .70). However, it is also plausible that the scales tap into different specific constructs. In fact, Farmer and Sundberg (1986), after observing a weak correlation between their newly developed BPS and the ZBS, proposed that the scales may assess different aspects of the boredom experience. Additionally, although the ZBS often demonstrates poor reliability, it is a commonly used measure of boredom, and thus, it is important to understand how it is similar or dissimilar to the BPS. Moreover, as evidenced by Vodanovich’s (2003) review, the ZBS has been shown to correlate with several variables of interest, including, for example, impulsivity (Eysenck & Zuckerman, 1978), extraversion (Eysenck & Zuckerman, 1978), narcissism (Emmons, 1981), drug use (Beauducel, Brocke, Strobel, & Strobel, 1999), and deviant behavior in school (Wasson, 1981). 1
Important differences between the scales are suggested not only from an empirical standpoint but also from a conceptual one. Specifically, in the construction of each scale, the authors appear to have defined boredom differently. According to Zuckerman (1979), the ZBS was developed to measure boredom as defined as “an aversion for repetitive experience of any kind, routine work, or dull and boring people and extreme restlessness under conditions when escape from constancy is impossible” (p. 103). Example items on this scale include “I have no patience with dull or boring persons” versus “I find something interesting in almost every person I talk to”; “I enjoy spending time in the familiar surroundings of home” versus “I get very restless if I have to stay around home for any length of time” and “There are some movies I enjoy seeing a second or even third time” versus “I can’t stand watching a movie that I’ve seen before.” Thus, boredom was conceptualized by Zuckerman as the tendency to be underaroused because of an impoverished or understimulating environment. The BPS, on the other hand, was developed to measure boredom in a way that emphasizes “one’s connectedness with one’s environment . . . as well as the ability to access adaptive resources and realize competencies” (Farmer & Sundberg, 1986, p. 10). Sample items include “Much of the time I just sit around doing nothing”; “I am often trapped in situations where I have to do meaningless things”; “It would be very hard for me to find a job that is exciting enough”; and “Time always seems to be passing slowly.” Thus, although some of the items on the BPS also highlight the role of the environment in boredom, the emphasis here is on the individual’s inability to become meaningfully connected to the external world rather than on the external world lacking in the stimulation she or he requires.
In sum, the BPS and ZBS are not highly correlated, and they were developed according to somewhat different definitions. Yet researchers show little awareness that they may not measure the same construct and freely generalize across studies using different scales—that is, they draw conclusions about boredom generally, without limiting the generalization of findings to the particular scale used. Moreover, although both the BPS and ZBS have been correlated with many constructs, they have not often been administered together within a single study (see Blaszczynski et al., 1990; Farmer & Sundberg, 1986; Vodanovich & Kass, 1990b, for exceptions), making it difficult to know which constructs are related to both scales and which constructs are uniquely associated with only one of the scales.
Purpose of the Present Research
The overarching purpose of the present research was to clarify the underlying construct that is measured by the BPS and ZBS. To this end, a series of studies employing correlation and regression analyses was carried out. Study 1 involved exploring whether postulated causes of boredom differentially predict scores on the BPS and ZBS. Study 2 involved determining whether the two scales are related to different emotional correlates of boredom. Study 3 examined the relationship between the two scales and the behavioral consequences of boredom.
Study 1
Causal Variables
Several causes of boredom have been proposed in the literature, and they can be organized into four distinct types: (a) emotional, (b) arousal, (c) cognitive, and (d) existential. These “theories” were referred to in order to establish relevant content domain with which to explore possible differences and areas of overlap between the two boredom scales, rather than to evaluate different models of boredom per se. The general prediction was that there would be significant differences between the scales and their relationships to the variables.
It has been suggested that lacking emotional awareness, or having the inability to identify, describe, and label one’s emotions; avoiding experience, or having the tendency to escape or avoid unwanted feelings and thoughts; and being sensitive to punishment may result in boredom. For example, several psychodynamic theorists have paid particular attention to the emotional antecedents of boredom. They have postulated that boredom is a state of wanting something to do but being unable to designate what it is that is desired because boredom signifies a repressed and threatening impulse (Fenichel, 1934/1953; Greenson, 1953; Lewinsky, 1943; Wangh, 1975). Boredom has also been theoretically linked to the tendency to seek arousal and reward. Some theorists have argued that individuals are driven to maintain an optimal level of arousal and that boredom ensues when there is a mismatch between this need for arousal and the availability of environmental stimulation (Berlyne, 1960; Csikszentmihalyi, 1975/2000; De Chenne, 1988; Hebb, 1966). Thus, variables related to arousal/reward seeking that might be expected to predict boredom include impulsivity and behavioral activation, or the tendency to respond to signals of reward by initiating goal-directed behavior. A third set of processes that have been viewed as a central cause of boredom are cognitive-attentional processes. For some theorists, boredom is the affective consequence one experiences when unable to focus or engage one’s attention (Fisher, 1998; Hamilton, 1983; Leary, Rogers, Canfield, & Coe, 1986). Thus, one might predict boredom to be related to attention deficits and associated cognitive errors. Finally, lacking life meaning and life satisfaction has also been connected to boredom, which is especially evident within existential theories of boredom (Bargdill, 2000; Frankl, 1984; Maddi, 1970). From this perspective, boredom is viewed as a symptom of a fundamental state of meaninglessness, suggesting a possible relationship between boredom and life meaning and life satisfaction.
Participants and Procedure
A total of 837 undergraduate students completed a battery of self-report scales that included both boredom scales as well as measures of the key causal variables outlined above. All participants received course credit for their participation. The mean age was 20.3 years (SD = 3.97, range 17-56 years), and 65% of the participants were women (n = 547). The sample was ethnically diverse: 37% identified as Caucasian, 18% as South Asian, 7% Chinese, 7% Arab/West Asian, 7% Black, 3% West Indian, 3% Biracial, 3% Filipino, 2% Aboriginal/First Nations, 2% Korean, 2% South East Asian, 2% Latin Canadian, and 6% as “Other.”
Measures
Boredom Scales
Boredom Proneness Scale
The BPS is a trait scale that measures “one’s proneness toward experiencing boredom” (Farmer & Sundberg, 1986, p. 5). The original BPS consisted of 28 true–false items, had an internal consistency alpha of .79, and a 1-week interval test–retest reliability of .83 (Farmer & Sundberg, 1986). Kass and Vodanovich (1990) converted the true–false format of the BPS into a 7-point Likert-type format to increase its sensitivity. The internal consistency alpha of the 7-point Likert-type version has been reported to range from .79 to .83 (Ahmed, 1990; McLeod & Vodanovich, 1991; Vodanovich & Kass, 1990a); and the test–retest reliability was reported to be .79 over a 1-week interval (Polly, Vodanovich, Watt, & Blanchard, 1993) and .76 over a 2-month interval (Fahlman et al., 2009). This 7-point version was used in all four studies. In the present study, coefficient alpha of the BPS was .83.
Boredom Susceptibility Scale
The ZBS is a subscale of Zuckerman’s (2005) Sensation Seeking Scale, Form V. Sensation seeking is a trait involving the “need for varied, novel, and complex sensations and experiences and the willingness to take physical and social risks for the sake of such experience” (Zuckerman, 1979, p.10). The ZBS, more specifically, measures one’s inability to tolerate monotonous environmental stimulation. The scale consists of 10 items arranged in a forced-choice format. Zuckerman, Eysenck, and Eysenck (1978) reported the internal consistency reliability of the ZBS to range from .56 to .65. A more recent review of 21 studies reporting coefficient alpha statistics for the ZBS found a mean and median coefficient alpha of .62 and .61, respectively (Deditius-Island & Caruso, 2002). Split-half reliability reports of the ZBS have ranged from .38 to .75 (Vodanovich, 2003), and the test–retest reliability after 3 weeks was reported to be .70 (unpublished data, cited in Zuckerman, 1979). Coefficient alpha of the ZBS in the present study was .57.
Scales Measuring Potential Causes of Boredom
Acceptance and Action Questionnaire (AAQ)
The AAQ contains nine items intended to measure experiential avoidance, or the tendency to escape or avoid unwanted feelings and thoughts. The internal consistency alpha of the scale was reported to be .70 (Hayes et al., 2004) and in the present study was .59.
Emotional Awareness Measure (EA)
The Emotional Awareness Measure is a 17-item measure of one’s ability to identify, describe, and label one’s emotions. The measure comprises items from three well-validated subscales, namely, the “difficulty identifying feelings” and “difficulty describing feelings” factors of the Toronto Alexithymia Scale (Bagby, Parker, & Taylor, 1994) and the “mood-labeling” factor of the Mood Awareness Scale (Swinkels & Giuliano, 1995). In a previous study, these three subscales were found to load highly onto a single factor, labeled “emotional awareness” (Eastwood et al., 2007). Coefficient alpha of this measure was .92 in the present study.
Sensitivity to Punishment and Sensitivity to Reward Questionnaire
This questionnaire contains two 24-item subscales, the Sensitivity to Punishment (SP) subscale and Sensitivity to Reward (SR) subscale (Torrubia, Avila, Molto, & Caseras, 2001). The SP scale assesses passive avoidance as well as cognitive processes such as worry in response to aversive consequences or novelty. The SR scale measures one’s tendency toward approach behavior in response to reward or nonpunishment. Internal consistency alphas have been reported to range from .82 to .83 for the SP scale and .75 to .78 for the SR scale (Torrubia et al., 2001). In the present study, coefficient alphas were .84. and .79, respectively
Adult ADHD Self-Rating Scale (ASRS)
The ASRS contains 18 items assessing symptoms of adult attention deficit hyperactivity disorder (ADHD) within the past 6 months (Kessler et al., 2005). The scale comprises two subscales: Inattention and Hyperactivity-Impulsivity. Coefficient alpha for the full scale has been reported to be .88 (Adler et al., 2006). In the present study, only the Inattention subscale (ASRS-Inatt) was used, which had a coefficient alpha of .73.
Attention-Related Cognitive Errors Scale (ARCES)
The ARCES comprises 12 items designed to measure one’s tendency to commit performance errors caused by attentional lapses (Cheyne, Carriere, & Smilek, 2006). The coefficient alpha for the scale was reported to be .89 (Cheyne et al., 2006). In the present study it was .86.
Barratt’s Impulsivity Scale-11 (BIS)
The BIS is a 30-item measure of the personality trait of impulsiveness (Patton, Stanford, & Barratt, 1995). Patton et al. identified three first-order factors of the scale: Motor Impulsiveness (BIS-Mot), Attentional Impulsiveness (BIS-Att), and Nonplanning Impulsiveness (BIS-NP). The BIS-Mot factor measures the drive to seek stimulation; the BIS-Att factor measures “focusing on the task at hand” and “thought insertions and racing thoughts” (p. 770); and the BIS-NP factor measures a tendency toward “planning and thinking carefully” and enjoyment of challenging mental tasks (p. 770). In the present study, coefficient alphas of these subscale were .84, .65, and .74, respectively.
Purpose in Life Test (PIL)
The PIL is a 28-item measure of the degree to which an individual experiences purpose in life (Crumbaugh & Maholick, 1964). Its internal consistency alpha has been reported to be .90 (Crumbaugh & Maholick, 1964) and in the present study was .84.
Satisfaction With Life Scale (SWLS)
The SWLS is a five-item scale measuring global life satisfaction (Diener, Emmons, Larsen, & Griffin, 1985). Coefficient alpha was reported to be .87 (Diener et al., 1985), which was identical in the present study.
Results
Several univariate outliers (i.e., greater than three standard deviations above or below the sample mean) were detected: eight on BPS, seven on ASRS, six on AAQ, three on BIS-Mot and PIL, and two on ARCES and BIS-NP; these scores were deleted for individual participants. All variables were approximately normally distributed. Two separate simultaneous multiple regression models were estimated, first regressing BPS on all variables (Model 1) and then regressing ZBS on all variables (Model 2; see Tables 1 and 2 for a summary of both models). A Bonferonni-corrected p level of < .025 (i.e., .05/2) was used to maintain family-wise alpha. The key question being addressed by these analyses was, What is the relative contribution of each variable, over and above other variables, in predicting a given boredom measure?
Study 1: Summary Table for Linear Regression Model Predicting BPS From Causal Variables.
Note. BPS = Boredom Proneness Scale; AAQ = Action and Acceptance Questionnaire; EA = Emotional Awareness Measure; SP = Sensitivity to Punishment Scale; ASRS-Inatt = Inattention Subscale of the Adult Attention Deficit Hyperactivity Disorder Self-Report Scale; ARCES = Attention-Related Cognitive Errors Scale; BIS-Att = Attentional Impulsiveness Subfactor of the Barratt’s Impulsivity Scale-11; BIS-NP = Nonplanning Impulsiveness Subfactor of the BIS-11; SR = sensitivity to reward; BIS-mot = Motor Impulsiveness Subfactor of the BIS-11; PIL = Purpose in Life Inventory; SWLS = Satisfaction With Life Scale; VIF = variance inflation factor.
Study 1: Summary Table for Linear Regression Model Predicting ZBS From Causal Variables.
Note. ZBS = Boredom Susceptibility Scale; AAQ = Action and Acceptance Questionnaire; EA = Emotional Awareness Measure; SP = Sensitivity to Punishment Scale; ASRS-Inatt = Inattention Subscale of the Adult Attention Deficit Hyperactivity Disorder Self-Report Scale; ARCES = Attention-Related Cognitive Errors Scale; BIS-Att = Attentional Impulsiveness Subfactor of the Barratt’s Impulsivity Scale-11; BIS-NP = Nonplanning Impulsiveness Subfactor of the BIS-11; SR = Sensitivity to Reward; BIS-Mot = Motor Impulsiveness Subfactor of the BIS-11; PIL = Purpose in Life Inventory; SWLS = Satisfaction With Life Scale; VIF = variance inflation factor.
Results indicated that the BPS and ZBS were distinct in a number of important ways with respect to the various causes of boredom. The BPS was positively associated with attention problems (BIS-Att and BIS-NP), emotional awareness, and purpose in life, whereas the ZBS was not uniquely related to these variables. On the other hand, the ZBS was positively associated with motor impulsiveness and sensitivity to reward and negatively associated with sensitivity to punishment, and the BPS was not uniquely related to these variables. Finally, the BPS was positively associated with experiential avoidance, whereas the ZBS was negatively associated with experiential avoidance. These findings suggest that the ZBS and BPS are differentially related to various proposed causes of boredom.
Study 2
Existing theories of boredom as well as previous qualitative and quantitative studies have suggested that boredom is related to various types of negative affect. For example, in qualitative studies boredom has been linked to anger (e.g., Bargdill, 2000), anxiety (e.g., Harris, 2000), and sadness or dysphoria (e.g., Martin, Sadlo, & Stew, 2006). Similarly, correlational studies have found significant associations between boredom and depression (e.g., Farmer & Sundberg, 1986), anxiety (e.g., Sommers & Vodanovich, 2000), anger (e.g., Rupp & Vodanovich, 1997), and neuroticism (e.g., Gordon, Wilkinson, McGown, & Jovanoska, 1997). The purpose of Study 2 was to examine the relationships between the ZBS and BPS and key emotional correlates of boredom in order to further explore and clarify the distinctness of these two measures.
Participants and Procedure
Participants were 233 undergraduate students, different from those in Studies 1 and 3. All participants in Study 2 completed a battery of self-report scales and received course credit for their participation. Mean age of the sample was 19.7 years (SD = 3.8, range 17-45 years), and 55% were women (n = 129). Participants identified with the following ethnic groups: 45% Caucasian, 14% South Asian, 10% Chinese, 9% Arab/West Asian, 6% Black, 6% Biracial, 4% South East Asian, 3% Korean, 1% West Indian, 1% Latin Canadian, and 3% identified as “Other.”
Measures
In addition to the BPS and ZBS, participants completed the following measures.
Center for Epidemiological Studies Depression Scale (CES-D)
The CES-D is a 20-item measure that assesses one’s current level of depressive symptomatology (Radloff, 1977). Radloff has reported the internal consistency alpha of the CES-D to be .85 and the test–retest reliability to range from .45 to .70. Coefficient alpha in the present study was .90.
Dysphoria Measure (Dys)
One aspect of depression that is of particular interest when understanding its relation to boredom is dysphoric mood. However, because the CES-D (like other depression measures) was developed to measure the full spectrum of depressive symptomatology, a specific measure of dysphoria was derived by reviewing existing mood (e.g., Profile of Mood States; McNair, Lorr, & Droppleman, 1971) and depression scales (e.g., Beck Depression Inventory; Beck, Steer, & Brown, 1996) and extracting relevant items. The measure contains eight items: “I feel down”; “I feel like crying”; “I feel blue”; “I feel downhearted”; “I feel sad”; “I feel unhappy”; “I feel gloomy”; and “I feel low.” Each is rated on a 7-point Likert-type scale, ranging from Strongly disagree to Strongly agree. In the present study, dysphoria scores were correlated with the CES-D at r = .76 and had a coefficient alpha of .96.
State-Trait Personality Inventory (STPI)
The STPI contains state and trait measures of depression, curiosity, anxiety, and anger (Spielberger, 1995). It consists of 80 items, with 10 items per subscale, all rated on a 4-point Likert-type scale. Because boredom was conceptualized as a trait by the authors of the BPS and ZBS, only the trait version of the scale was used in the present study. Specifically, the trait measures of anxiety (STPI-Anx) and anger (STPI-Ang) were included. Coefficient alphas for these trait subscales have been reported to range from .88 to .92 for anxiety and from .88 to .92 for anger (Spielberger, 1995). In the present study, coefficient alphas for the STPI-Anx and STPI-Ang were .66 and .81, respectively.
Big Five Inventory–Neuroticism Subscale (BFI-N)
The BFI-N contains eight items measuring the personality trait of neuroticism, which “contrasts emotional stability with a broad range of negative affects, including anxiety, sadness, irritability, and nervous tension” (Benet-Martinez & John, 1998, p. 730). Coefficient alpha of the subscale has been reported to range from .80 to .84 (Benet-Martinez & John, 1998) and in the present study was .84.
Results
One univariate outlier was detected on the CES-D; this score was deleted. All variables were approximately normally distributed. Zero-order correlations were examined for differences between the BPS and ZBS and key correlates (see Table 3 for correlations and coefficient alphas for all scales). A Bonferonni-corrected p level of <.005 (i.e., .05/11) was used to maintain family-wise alpha. Consistent with past findings, the ZBS and BPS were weakly correlated (r = .21). Moreover, the BPS was positively correlated with all negative affective measures, including depression (r = .52), dysphoria (r = .45), anxiety (r = .51), and anger (r = .36), as well as with neuroticism (r = .40), suggesting that higher levels of boredom proneness are associated with higher levels of each of these constructs. The ZBS, by contrast, was only weakly related to anger (r = .13), weakly and negatively correlated with neuroticism (r = −.17), and did not correlate with any of the other emotion or personality measures. These findings suggest that the ZBS and BPS are differentially related to emotional correlates of boredom.
Study 2: Zero-Order Correlations (N = 233).
Note. BPS = Boredom Proneness Scale; ZBS = Boredom Susceptibility Scale; CESD = Center for Epidemiological Studies Depression Scale;Dys = Dysphoria Measure; STPI-Anx = State-Trait Personality Inventory–Anxiety Scale; STPI-Ang = State-Trait Personality Inventory–Anger Scale; BFI-N = Big Five Inventory–Neuroticism Subscale. p < .005 for correlations in boldface.
Study 3
The aim of Study 3 was to explore differences between the BPS and ZBS by examining their relationships with several maladaptive behaviors that have been previously linked to boredom. For example, prior qualitative and quantitative research has linked boredom to addictive behaviors such as binge eating (e.g., Stickney & Miltenberger, 1999), gambling (e.g., Mercer & Eastwood, 2010), and alcohol and drug use (e.g., Iso-Ahola & Crowley, 1991; Paulson et al., 1990). For each of these behaviors, boredom is presumed to be an antecedent, leading one to engage in maladaptive efforts to overcome one’s boredom.
Participants and Procedure
A total of 178 undergraduate students participated in Study 3 for course credit. All participants completed a battery of self-report measures, including both boredom scales and self-report measures of the maladaptive behaviors listed above. The sample had a mean age of 19.9 years (SD = 3.01, range 18-47 years), and 51% were women (n = 90). Participants were ethnically diverse: 37% identified as White/Caucasian, 16% as South Asian, 12% Chinese, 9% Arab/West Asian, 7% Black, 6% Biracial, 5% South East Asian, 3% West Indian, 2% Filipino, 2% Latin Canadian, 1% Korean, 1% Aboriginal/First Nations, and 1% as “Other.”
Measures
In addition to the BPS and ZBS, participants completed the following measures.
Emotional Eating Scale (EES)
The EES is a 25-item measure that assesses the degree to which one experiences an urge to eat in response to positive and negative emotion (Arnow, Kennedy, & Agras, 1995). Coefficient alpha for the total scale was reported to be .81 (Arnow et al., 1995) and in the present study was .95.
South Oaks Gambling Screen (SOGS)
The SOGS consists of 20 yes/no items based on the Diagnostic and Statistical Manual of Mental Disorders–Third Edition (DSM-III; American Psychiatric Association, 1980) criteria of pathological gambling and was designed to screen for cases of probable pathological gambling (Leiseur & Blume, 1987). Scores indicate a specific category of gambling status (i.e., “no problems with gambling,” “some problems with gambling,” or “probable pathological gambler”). In this study, categories were not used, mainly because our sample was randomly drawn from a population of undergraduate students, and instead, total scores were used to measure degree of gambling problems along a continuum. The coefficient alpha of the SOGS has been reported to be .81 (Ferris & Wynne, 2001), and the test–retest reliability to be .75 (Ferris & Wynne, 2001). In the present study, coefficient alpha was .82.
The Short Michigan Alcohol Screening Test (SMAST)
The 10-item SMAST (Selzer, Vinokur, & van Rooijan, 1975) is a shortened version of the 25-item Michigan Alcohol Screening Test (MAST; Selzer, 1971), which is a self-report questionnaire designed to detect symptoms of problematic drinking as well as associated negative consequences. In the present study, the SMAST scores were calculated on a continuum with higher scores indicating greater alcohol problems. Coefficient alpha of the SMAST in the present study was .67.
Results
Two univariate outliers were detected on the EES; these scores were deleted. Multiple regression was used to determine whether the two boredom scales were uniquely and differentially related to the outcome variables. Specifically, three linear regression models were estimated in which the ZBS and BPS were predictors of SMAST scores (Model 1), log-transformed SOGS scores (Model 2), and EES scores (Model 3). A Bonferonni-corrected p level of <.02 (i.e., .05/3) was used to maintain family-wise alpha. Table 4 presents the results of these analyses. In an initial regression analysis, the distribution of residuals for SOGS scores was positively skewed and thus this variable was transformed using a logarithmic function.
Study 3: Summary Table for Linear Regression Models Predicting Alcohol Use, Problem Gambling, and Emotional Eating.
Note. BPS = Boredom Proneness Scale; ZBS = Boredom Susceptibility Scale; SMAST = Short Michigan Alcohol Screening Test; SOGS = South Oaks Gambling Screen; log-SOGS = log-transformed SOGS; EES = Emotional Eating Scale.
Results indicated that the ZBS was significantly and uniquely related to alcohol problems and gambling behavior, such that higher ZBS scores were associated with higher levels of both of these behaviors. However, the ZBS was not uniquely predictive of emotional eating. In contrast, the BPS was not uniquely related to either alcohol problems or gambling problems but was uniquely predictive of emotional eating, such that higher BPS scores were associated with higher levels of emotional eating. These findings suggest that the ZBS and BPS are differentially related to various pathological consequences of boredom.
General Discussion
Previous research has clearly demonstrated that boredom is an important construct that deserves empirical consideration. Yet historically boredom has not received sustained or systematic attention, in part because existing self-report measures of boredom are limited both theoretically and psychometrically (see Vodanovich, 2003).
The current findings revealed that the BPS and ZBS, two commonly used measures of trait boredom, do not measure the same underlying construct. The variables that most clearly discriminated between the two scales were neuroticism and experiential avoidance. Whereas the BPS was associated with higher levels of neuroticism and experiential avoidance, the ZBS was associated with lower levels of these variables. In addition, there were several variables that were uniquely associated or correlated with one scale and not the other. Specifically, unlike the ZBS, the BPS was uniquely associated with increased attentional and nonplanning impulsivity and decreased emotional awareness; it was correlated with anxiety, depression, and dysphoria; and it was uniquely predictive of emotional eating. On the other hand, the ZBS, and not the BPS, was uniquely associated with increased motor impulsivity, increased sensitivity to reward, decreased sensitivity to punishment, and increased problematic gambling behavior and alcohol use.
Thus, one way of summarizing the distinction between these scales is that boredom as measured by the BPS appears to be characterized by negative emotionality and withdrawal from the world, and boredom as measured by the ZBS seems to be characterized by the opposite, a moving forward toward the world in attempt to seek stimulation or arousal. Indeed, this conceptualization may help to explain the different relationships observed between the two scales and the emotional correlates and behavioral consequences. Whereas individuals with high ZBS scores may be more likely to be motorically impulsive and engage in risky or excitement-seeking behaviors (e.g., alcohol abuse or problem gambling behaviors) in order to obtain reward, experience positive affect, and elevate their level of arousal, those with high BPS scores appear more likely to engage in more socially isolated behaviors (e.g., emotional eating) to avoid punishment and to experience affective disturbances (e.g., anxiety or depression).
A conceptual link can also be made between the two scales and the broad “internalizing” and “externalizing” classification of disorders. Numerous factor analytic studies have converged on these two primary dimensions of psychopathology, one representing disorders involving an inward expression of emotional distress (i.e., internalized) and the other representing disorders involving an outward expression of emotional distress (i.e., externalized; see Krueger, 1999; Krueger, McGue, & Iacono, 2001; Krueger & Tackett, 2003). For example, anxiety and depression are presumed to reflect internalizing problems, and substance abuse is presumed to reflect externalizing problems. Based on the present findings and differences found between the two boredom measures, it seems that the BPS is more closely related to internalizing disorders and the ZBS to externalizing disorders.
Still another framework that can be used to make sense of the present findings is Clark’s (2005) tripartite model of personality-psychopathology relations. According to this model, three innate temperament dimensions—negative affectivity, positive affectivity, and disinhibition (vs. constraint)—underlie most, if not all, psychological disorders. Indeed, adult personality traits are posited to emerge through differentiation from these temperaments, and, when at extremes, particular temperament-personality dimensions are considered risk factors for specific disorders. For example, research in this area has found that negative emotionality is associated with a broad range of psychopathology, including depression, anxiety, substance dependence, and conduct disorders (Krueger, Caspi, Moffitt, Silva, & McGee, 1996). Positive temperament has been negatively linked to psychological disorders such as depression, schizophrenia, and social phobia (Mineka, Watson, & Clark, 1998). Finally, the disinhibition (vs. constraint) dimension is primarily associated with externalizing disorders (e.g., substance abuse and antisocial personality disorder or conduct disorder; Kendler, Prescott, Myers, & Neale, 2003; Krueger et al., 1996; Lynam, Leukefeld, & Clayton, 2003). Present findings suggest that boredom as measured by the BPS is more closely related to higher levels of negative affect, lower levels of positive affect, and high levels of self-constraint. The ZBS, alternatively, seems to be related to high levels of positive affect and, more critically, to low levels of self-constraint.
Taken together, the present findings clearly support the notion that the BPS and ZBS do not measure the same experience or construct. However, a critical question is raised by these findings: Do these scales both measure the experience of boredom but just different types or aspects of it? Or, does only one scale measure boredom and the other not measure it at all? It may be that the ZBS is merely a measure of sensation seeking and that the BPS taps into boredom, or, alternatively, perhaps the BPS simply assesses neuroticism, or the tendency to experience any negative affect, and the ZBS more accurately reflects the concept of boredom. This is an important and unresolved theoretical question regarding the definition of boredom. Findings from the present study do still suggest, however, that the two constructs measured by the BPS and ZBS are related and contain some overlap. Across the three studies, BPS and ZBS had an average correlation of .3, indicating they are related to each other in some manner. Furthermore, in Study 1, lacking life meaning uniquely predicted both the BPS and ZPS. Finally, in Study 2, anger was positively correlated with both the BPS and ZBS. These areas of overlap suggest that the scales do not measure something entirely different.
Limitations of the Present Research and Directions for the Future
The present research contains several findings that support a broadened understanding of boredom and its measurement; however, there are several limitations. First, all studies were conducted with a young and relatively educated adult sample. It is plausible that the characteristics of this sample may have influenced the results in some important way. For example, perhaps adolescents or young adults become more easily bored by routine and repetition or by a lack of environmental stimulation (i.e., the type of boredom measured by the ZBS) than middle-aged or older adults (e.g., Essed et al., 2006). It is possible that observed relationships between boredom, or different types of boredom, and other constructs (e.g., depression, inattention, alcohol abuse) differ in populations with other demographic characteristics. Additionally, although there has been some cross-sectional work demonstrating important developmental periods in which people are prone to boredom (e.g., adolescence, old age; Hamilton, 1983), there has not been any systematic empirical investigation into variation in boredom levels, or types, as a function of age.
It is also possible that different relationships would have been observed in a clinical sample. For instance, McCormick (1988) proposed that problem gamblers can be divided into two distinct subtypes: those who gamble to alleviate a negative mood state and those who gamble to alleviate a state of underarousal. In the former group, the depressed pathological gambler is relieved of dysphoric mood by the affect-enhancing excitement produced by gambling. Alternatively, the chronically understimulated gambler, who experiences low frustration tolerance and a need for varied stimulation and constant arousal, characterizes the latter group. The arousal produced by gambling in this latter group acts as negative reinforcement alleviating under stimulation, thereby reinforcing further gambling behavior. Recent empirical evidence supports this distinction (Vachon & Bagby, 2009). Using cluster analysis on personality traits with a sample of gamblers, Vachon and Bagby identified three types of problem gamblers: a “hedonic” type (characterized by excitement seeking and positive affect), a “demoralized” type (characterized by negative affect, low positive emotionality, and disinhibition), and a “simple” type (characterized by low comorbid psycholopathology and personality traits near the normative mean). Given the present findings suggesting different types of boredom, one related to negative affect (i.e., BPS) and one to underarousal (i.e., ZBS), research on the full range of gambling problems might find evidence for two types of boredom that are consistent with the subtypes of problem gambling proposed by McCormick (1988) and Vachon and Bagby (2009).
Moreover, although the present work demonstrated that gambling was related to the ZBS and not the BPS in an undergraduate sample, research conducted by Blaszczynski et al. (1990) with pathological gamblers demonstrated the opposite—namely, that problem gambling was related to the BPS and not the ZBS. Perhaps the relationship between boredom and problem gambling is best understood in two phases, where boredom as measured by the ZBS plays a more important role in the development of gambling problems and boredom as measured by the BPS in the maintenance. Future research with this and other clinical populations might help further clarify the exact role of boredom in these contexts.
The present research has established that the BPS and ZBS do not measure the same underlying construct. This finding provides an important foundation and constraint for subsequent empirical investigations of boredom and also subsequent theoretical work attempting to understand and define boredom. Indeed, as mentioned earlier, the present findings raise important theoretical questions. Specifically, the present findings could be interpreted to mean that there may be two kinds of boredom and that each is related to different causal, correlational, and outcome variables. Alternatively, the present findings could be interpreted to mean that only one of the BPS or ZBS is actually measuring boredom. This second possibility could only be evaluated with a strong, mutually agreed-on definition of boredom in hand. Unfortunately, such a definition does not yet exist (Vodanovich, 2003). However, future theoretical accounts of boredom must be able to accommodate the present empirical findings; thus, our work has set the stage for important theoretical work. Furthermore, future empirical investigations of the various theories of boredom can only be carried out once we are clear on what our self-report scales measure, and the present findings contribute to this clarity. In sum, the present findings provide important constraints and guidelines for subsequent theory-informed investigations that will not only help broaden our understanding of boredom as a construct but also inform the development of a more unified and comprehensive definition of boredom, as well as a psychometrically sound measure of boredom.
Finally, given the psychometric limitations (e.g., poor reliability) of the ZBS, the present study needs to be replicated using a more psychometrically sound measure of the type of boredom measured by the ZBS. Similarly, poor reliability demonstrated by other measures in the present study, including the STPI and the AAQ, suggests that future researchers use other, more sound, measures of the constructs measured by these scales.
Implications and Conclusions
The present findings have critical implications for the interpretation of past work on boredom and for future research. One important implication is that researchers should not generalize across studies without thoughtful consideration of the differences between boredom scales. Moreover, researchers should be more precise regarding the concept of boredom and how it is operationalized. In fact, based on the present results it is clear that researchers need to be aware that the BPS and ZBS tap different aspects of boredom. However, without specificity regarding the type of boredom that is being investigated, perhaps researchers should consider use of both scales to further our understanding of boredom. In fact, it may also be helpful to include measures of neuroticism, sensitivity to reward and punishment, and sensation seeking in future studies on boredom, both to further clarify the construct of boredom and to establish its incremental validity. It is possible that the concept of trait boredom is redundant with the existing concepts of neuroticism, sensitivity to reward and punishment, and sensation seeking.
Considering the close relationship between boredom and psychopathology, it is clear that the experience of boredom is a significant psychological construct. The present results further highlight the importance of studying boredom in its own right and in relation to emotional, personality, and behavioral constructs. Findings from the present work also offer some conceptual clarity regarding the concept of boredom and its measurement. A broad conclusion that can be drawn is that existing boredom measures, namely, the BPS and ZBS, do not converge to measure the same construct. Instead, these scales appear to measure a different type or aspect of boredom. Whereas boredom measured by the BPS appears to be characterized by experiential avoidance, neuroticism, and negative affect generally, boredom as measured by the ZBS seems to be characterized by increased sensitivity to reward and lowered sensitivity to punishment and neuroticism. Importantly, distinguishing between these self-report scales contributes to our understanding of the boredom more broadly and can move us toward a comprehensive, well-informed, and universally agreed-on definition of boredom.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Partial support for this research was provided by the Ontario Problem Gambling Research Centre (OPGRC).
